The votive Column from the Roman Temple Precinct at Springhead

THE VOTIVE COLUMN FROM THE ROMAN TEMPLE PRECINCT AT SPRINGHEAD T. F. C. BLAGG, M.A. The excavations at Springhead produced fragments of two pieces of Roman decorated stonework which are of particular interest: a Corinthian column capital;' and a column shaft decorated with overlapping leaves.2 The purpose of this paper is, first, to consider the dating and the possible antecedents of the former; secondly, to reconsider the interpretation of the latter; and so to argue that both formed part of a free-standing votive column, probably erected in the second half of the second century or early in the third. The Corinthian Capital A pit dug in front of a brick pedestal 1.31 m. square which stood in the temple precinct at Springhead contained fragments of a Corinthian column capital, of which four were published: two joining pieces from the upper of the two tiers of acanthus leaves from the bottom half of the capital; a piece from the lower tier; and the corner of the abacus and volutes at the top. The other fragments now in Gravesend Museum presumably came from the same pit, which is dated to or after the late third century by a coin of Postumus. The capital was carved frorri Jurassic limestone.3 It is unlike any other found in Britain.4 1 W. S. Penn, 'The Romano-British Settlement at Springhead. Excavation of the Watling Street, Shop and Pedestal, Site B', Arch. Cant., lxxii (1958), 87, 95, Fig. 11 and 108-110 (Nos. 2 A-C); hereafter cited as Penn (1958). 2 W. S. Penn, 'Springhead. Temple VI/Gateway', Arch. Cam., lxxxii (1967), 111, I 13 Fig. 4 and 123, Nos. 13 and 14; hereafter cited as Penn (1967). 3 Penn (1958), 107, Appendix 2. 4 For a brief preliminary account of Romano-British Corinthian capitals, and for the descriptive terminology, see T. F. C. Blagg, 'Schools of Stonemasons in Roman Britain', in J. Munby and M. Henig (edd.), Roman Life and Art in Britain, British Archaeological Reports 41, 1977, 61-67, Figs. 4.4 and 4.5 and pls. 4.11-4.IV. 223 T.F.C. BLAGG The decoration is well-preserved. The abacus is carved with slightly inclined fluting, and has a deeply-cut V-shaped recess at the end. The volutes, with their spiral terminals linked by a cross-bar, rise well above the leaves of the calyx. The acanthus tiers have leaves with a very slight overhang of the central lobe, and prominent middle lobes, each with three teeth. The central tooth is broad and has an ogival outer edge, terminating in a point. The lower lobes also have three teeth, the uppermost of which overlaps the middle lobe. The centres of the lobes are deeply scooped out to form a spoon-shaped channel within a flat border round the edge of the leaf. In the upper tier, a long downwards-pointing triangle is cut below the join between the middle lobes and the mid-rib; in the lower tier, this is replaced by three curved incisions. The mid-rib is outlined by deep grooves, and tapers towards the top. The capital appears to have been carved in one piece, terminating below the astragalus (the moulding between the capital proper and the column shaft), part of which is preserved on one of the unpublished fragments. The leaves are well-spaced, with the mid-ribs of those of the upper tier running down to the astragalus. Not enough pieces survive to permit a precise reconstruction of the overall dimensions, but the upper tier measured about 56 cm. across the top of the middle lobes. Since the leaves are only slightly splayed outwards, the diameter of the capital at the shaft can be estimated as about 50 cm. Its height was in the region of 65 cm. or a little more, the height of the abacus being 6.5 cm. and that of the two acanthus tiers 35 cm. The excavator's estimate of 29 in. (73.5 cm.) is, if anything, a· little on the high side, based as it is on the proportions of capitals from Rome itself. Provincial masons did not necessarily follow these too closely. There is no evidence that the capital had a fluted column, as was attached to it in the original reconstruction drawing.5 To the contrary, it will be suggested below that a column of a different kind can be attributed to it. It seems likely that the column stood on the pedestal at the bottom of which the fragments of the capital were found, as indeed the excavator suggested. The construction of the pedestal is dated in the report to the early second century, on the evidence of coarse pottery in the layer below the chalk make-up associated with the pedestal.6 The quantity of pottery is not stated, and in any case it provides only a terminus post quern for the event. Furthermore, there are grounds for thinking that the published dating of the pottery at s Penn ( 1958), Fig. 11, no. 3. The scale of the reconstruction is 1/32, not 1/40 as stated. In other respects than the fluting, the drawings reproduce the appearance of the fragments most faithfully. 6 Penn (1958), 83-85. 224 THE VOTIVE COLUMN AT SPRINGHEAD Springhead is a little too early.7 A later archaeological date for the pedestal would harmonize much better with the date one would prefer for the capital on stylistic grounds, but so long as the relative order is maintained, the gap in time is not of great concern. The sharply outlined spoon-shaped channels in the leaves are the first indicators of a date no earlier than the middle of the second century. The parallels for this feature are not to be found in Britain, but in eastern Gaul and the Rhineland. The Corinthian capitals from the latter area, with comparanda cited from further afield, formed the subject of a monograph by Kahler, which is the only detailed treatment of this aspect of Roman architectural ornament in the western provinces.8 The spoon-shaped channels are not found on capitals of Ka.Wer's first- and early second-century Form C, from which most of the Corinthian capitals in the south of Britain are derived, but seem to be foreshadowed on some of those of his Form H, notably that on the Iphigenia monument at Neumagen, dated by von Massow to the middle and by Kahler to the second half of the second century.9 They occur also on later monuments at Neumagen, together with the ogivally-shaped lobes to the leaves. 10 Closer to the way in which the leaves of the Springhead capital are scooped out are Ka.bier's J 6 (Mainz) and K 6 and K 7 (Strasbourg), all of which he dates to the third century. 11 The K 6 ea pita! also has the pointed ogival outline to the leaf and the incisions below the cavities where the lobes join, though in other respects the disposition of the leaves of these capitals is not similar; indeed, their carving is considerably less accomplished than that of the Springhead capital. The dating of these pieces is stylistic, but is based on comparison with the appreciable number of Kiihler's Form J to M capitals from Jupiter columns, dated by their inscriptions to the late second or third centuries. The chronology therefore seems reasonably sound. Further examples assist in defining the area in which these stylistic features occur. The fragment of an abacus and volutes from Gerouville in Belgian Luxembourg has an almost identical leaf to that of Springhead, and the volute is similarly treated, though it lacks the transverse bar and the abacus is decorated with leaves, not fluting. 12 7 I am grateful to Mr. A. P. Detsicas and Mr. S. R. Harker, for their information on this point. Mr. Detsicas comments that the relevant layer (Stratum C in the excavation re.r,rt) contains sherds datable to A.O. 80-200. H. Kahler, Die Romischen Kapitelle des Rheingebiets (Berlin 1939). 9 W. von Massow, Die Grabmiiler von Neumagen (Berlin and Leipzig 1932), 51 ff. and 280-285; Kahler, op. cit., 49, no. H 34, and 51-52. 10 Von Massow, op. cit., no. 184 and Tafn. 32-34. 11 Kahler, op. cit., 65-66. 12 Musee du Luxembourg, Pierres scu/ptees et Inscriptions de /'Epoque romaine, Catalogue par E. Wilhelm (Luxembourg 1974), 65 and 102, no. 419. 225 T.F.C. BLAGG. The scooped leaf with ogival tip also appears on the capital of a small imbricated column from Brumath, Alsace, discovered with thirdcentury material which provides a terminus ante quem.13 It occurs again, in Lorraine, on the figured capital of a Jupiter column from Merten, though on that piece the side lobes of adjacent leaves are replaced by a single small leaf between them. 14 The same scooped lobes appear, this time with semicircular nicks under the cavities between the middle lobes and the mid-ribs, on the capital of an engaged halfcolumn built into a late Roman wall at Dijon. 15 Again, the single small leaf replaces adjacent side lobes; the leaves are clumsier and the teeth more elongated. Otherwise, it provides the closest parallel for the detail of the Springhead capital. In all of these further examples, the placing of the leaves close together, and also the single small leaf on the two capitals last cited, as also on Kii.hler's capital J 6 from Mainz, are features which suggest that they are stylistically later than the capital from Springhead. Unfortunately, none is dated archaeologically. Considering the earlier evidence, however, it would seem likely that the Springhead capital should be placed some time in the second half of the second century or early in the third. It is unique in Britain, and its decoration cannot be shown to have developed from earlier Romano-British examples. It should thus probably be assigned to a mason from eastern Gaul, working in an earlier version of the style which is represented by the more evolved examples cited above. The Column Shaft Seven pieces of decorated column shaft have been found at various points in the Romano-Celtic temple complex. All are identical in technique. The decoration consists of imbricated leaves, that is, leaves which overlap like the tiles of a roof or the scales of a fish. The largest piece was part of a half-drum; a rectangular projection from the rear, 6 cm. deep, was rebated 4 cm. back from the face. This suggested to W. S. Penn, the excavator, that the half-column had been engaged to a wall. This and another piece were published as having been found within a square building, Temple VI/Gateway, considered to have been the entrance to the precinct. 16 One (it is not stated which) was found in the rubble filling of a pit dug in front of the altar which stood in the centre of the structure, the other by the side of the steps. A third, 13 Gallia, xxviii (1970), 337 and 338, Figs. 34 and 35. 14 E. Esperandieu, Receuil general des Bas Reliefs, Statues et Busies de la Gaule romaine (paris 1907-1966), v, 4425 and plate on p. 452. 15 Ibid., ix, 7188. 16 Penn (1967), 114-115. 226 THE VOTIVE COLUMN AT SPRINGHEAD unpublished, piece was found 'a• few yards away from the building, together with a number of carved stone fragments', and a fourth piece exhibited with the others in Gravesend Museum may perhaps be identified as one of those carved fragments. Penn interpreted his building as being open at the front and back, with columns flanking the street gate and steps at the rear going down into the precinct, giving 'the appearance of a single-arched triumphal arch'. This might seem straightforward, but there are problems. ( 1) Penn 's description must be corrected in some particulars. The pieces are incorrectly identified in the report, not as column shafts but as Corinthian capitals. They are carved from oolitic limestone, not millstone grit, as stated. They are said to be of different sizes. It is true that the width of the leaves varies, from 13 to I 4.5 cm. If the curvature of some fragments seems to be slightly greater, this is not easy to express precisely, since they are all small, and the wavy outline of the leaves produces perimeters which do not confonn exactly to a true circle. In any case, such difference in diameter as there may be is explicable as that between higher and lower drums in a single tapering column shaft. It is not therefore necessary to suppose that there were two columns of different dimensions. It seems that the wrong identification as Corinthian capitals misled Penn about the significance of the variations. (2) Only the foundations of the Temple/Gateway's walls survived, except possibly on the street side, the nature of which is not made entirely clear. There is no positive evidence, such as a stylobate block or a projecting plinth, for the placing of engaged columns in the positions suggested, or in any other position in the building. (3) Since that excavation, three further fragments with imbricated leaves have been found, in rubble deposits further north-west in the temenos, near Temples IV and VII, 17 that is, on the opposite side of the pedestal next to which the pieces of the Corinthian capital were found. The reasons for supposing that the column formed part of the Temple VI/Gateway building are thus considerably weakened. It is also difficult to find a suitable home for it in either of the two main temples on the site (I and II). 18 Neither of these had an external portico. Although it is possible that the antae added in front of them had their roofs supported by columns, the walls of Temple I's antae are only 35-38 cm. wide, and those of Temple II about 45 cm. In neither case are they large enough to have carried a column of the size of that to which these fragments belonged, which was about 52 cm. in 17 Information from S. R. Harker. 18 W. S. Penn, 'The Romano-British Settlement at Springhead; Excavation of Temple I, Site Cl', Arch. Cant., lxxiii (1959), 1-61; 'Springhead: Temples II and V', Arch. Cant., lxxvii (1962), 110-132. 227 T.F.C. BLAGG diameter, and which would have had a base with mouldings of even greater diameter. The only possible places in any of the excavated buildings where the foundations for a column of this size exist are the nine brick plinths, c. 75 cm. square, which carried the internal supports of Temple II. There is no particular reason to suppose that these plinths carried columns rather than rectangular piers of brick. To do so would seem to demand an unduly lavish use of limestone, considering the distance it had to be brought and the otherwise relatively simple character of the building and, though not refutable, seems implausible. A further possibility remains, however. It depends, as in fact the possibility just considered would also depend, on accepting that in spite of appearances the pieces came from a whole column and not from an engaged half column, or columns, and that in the case of the piece with the rectangular projection it had been necessary to join two halves vertically to form a complete drum, a need explicable if the blocks of limestone, transported from some distance, were not large enough to form a complete drum in every case. Against this suggestion is the observation that the technique does not seem to have been used elsewhere in Britain. In favour of it is the fact that the decoration does not end neatly at the edge of one leaf or at the midrib of the leaf below, as one might expect if it had been laid out in relation to an integral half-column, but is carved as if it continued on to an adjoining stone. Conclusion If the suggestion just made is accepted, it would seem highly likely that the shaft was that of a votive column, and that it stood on the pedestal at the bottom of which the pieces of Corinthian capital were found. Although no fragment of column shaft was found in the same place, others have been found in rubble layers in the precinct a short distance away to both east and west. These layers presumably derive from the demolition of the column and perhaps of other buildings in the complex. The frequency with which votive columns were decorated with imbricated leaves is well established. 19 The size is right: the shaft was about 52 cm. in diameter, and the capital is estimated to have been 50 cm. in diameter at the bottom; there is room for an adjustment of a couple of centimetres to both estimates. While the stone of which the column shaft was made has not been examined petrologically, it is certainly of oolitic limestone, and looks to be the same as that of the capital. The hypothesis is economical, even though the evidence which supports it is largely circumstantial. 19 H. Walter, La Colonne ciselee dans la Gaule romaine, Annales litteraires de /'Vniversite de Besan􀀟on, 21 (Paris 1970). 228 THE VOTIVE COLUMN AT SPRINGHEAD This conclusion follow's Peon's original insight, that the pedestal had supported a free-standing column, and that the Corinthian capital had belonged to that column. The reinterpretation of the significance of the pieces of imbricated column shaft may indeed be thought to strengthen Penn's hypothesis, which was advanced, of course, several years before the pieces of shaft were found. While he cited the Column of Phocas in the Forum at Rome by way of comparison, a parallel closer to home would be the Jupiter columns which are a common feature of the Rhineland and north-eastern Gaul, the area in which the stylistic antecedents of the capital also lie. Such columns usually had a shaft with imbricated leaves and Corinthian capitals, and were surmounted by statuary .20 A number of examples has been identified in Britain.21 Since no appropriate inscription or statuary has been found at Springhead, a positive identification of its column as dedicated to Jupiter would go beyond the available evidence, however attractive a possibility it may seem. 20 Walter, op. cit.; G. Bauchhenss, Die Juppitergigantensiiule (Stuttgart 1976). 21 Inscribed pedestals from Chichester and Cirencester: RIB 89 and 103; half of an octagonal pedestal with busts of the Gods of the week, another common feature of Rhineland Jupiter columns, from Great Chesterford, Essex: VC H (Essex) iii, pl. xviii, c; and a similar octagonal pedestal with niched figures, found in loose association with part of a Corinthian capital at Irchester: P. Woodfield, 'Roman architectural Masonry from Northamptonshire'. Northants. Archaeol., xiii (1978), 68-69, 81-82 and Fig. 6. Columns with imbricated leaves have been found at Caerwent, Catterick, Cirencester, Corbridge, Gloucester, Leicester, London, Ringstead (Northamptonshire), Winchester, Wroxeter and York. Two of these, Catterick (JRS, I (1960), 218) and Gloucester (Museum no. A.2648, unpublished, had Corinthian capitals. They and most of the others may be thought likely to have come from votive columns, though not all: the Caerwent example appears to have formed the internal order of the Forum basilica (Archaeologia, lxi (1909), 572). 229